EA and their Football Titles - Thoughts on game design

[lilblaze]

Can a moderator change the title to "EA and their Football Titles!" Please?

First off I am a college student studying to become a Software Engineer. (Makes video games)

I picked up a book called Fundamentals of Game Design and began reading. As I approached the Sports Games sections, I specifically looked for a football section... Come to find out, it's pretty much the basis of the whole chapter.

Here are some key things to note about EA and their football development team:

1. Madden is NOT REDESIGNED every year, nor is its code rewritten. Electronic Arts updates its artwork, video sequences and adds new features every year; only receiving a major overhaul every 4 to 5 years. Typically when a new console is launched. The majority of the design work each year consists of TUNING and IMPROVING the gameplay and adding MORE FEATURES.

2. In theory, the best, current football game available on the market today should be NCAA Football 12 and All-Pro Football 2k8. The reason I say this is because the largest, single design task that goes into DEVELOPING MADDEN is RESEARCH!

3. Players greatly enjoy setting up their own strategies, adjusting how the athletes will behave and what roles they will play in a team sport. This can be replicated following in 2k Basketball's foot steps. (Very similar to the Deny Ball/Play Tight defensive role settings. This can be implemented in Madden under a Coach Settings menu; accessible when the game is paused.)

4. During play, your game will be running a physics engine that DETERMINES THE BEHAVIOR OF MOVING BODIES in the match. (Think Psychic DB'S IS JUST BAD CODING? It's a failing Game Engine.. Frostbite 2 anyone???)

5. Even in a highly realistic game, YOUR OBJECTIVE IS TO PROVIDE AN ENJOYABLE EXPERIENCE, not a matchstick simulation of nature. (Madden 12 has become very repetitive, NCAA 12 has become very bland. There's just never any exciting plays.)

6. The ATHLETES in a sports game MUST BEHAVE LIKE HUMANS, and THAT MEANS DELIBERATE, INTELLIGENT ACTION.

7. Athletes must respond intelligently to a number of possible events. A HIGHLY COMPLEX SIMULATION SUCH AS AMERICAN FOOTBALL, with 22 players on the field at a time, presents the opportunity for hundreds of events. (Everything that happens on the field has an IF & THEN effect. This led me to believe that the task is a lot more difficult than I initially thought. There are small tweaks that are necessary so that flaws can be eliminated.)

8. Gameplay Code should be designed like a flowchart.

All Code in relation to gameplay, needs an (IF code) that trigger animations. (THEN codes) Should follow and trigger the next animation.
EXAMPLE: Defensive Code
Look for the ball > In position > Go for swat, Interception, Tackle (Eliminates the EYES IN THE BACK OF THE HEAD BALL SWAT)
Look for the ball > Not In position > Look for Ball Carrier > Sprint towards Ball Carrier > Tackle (Stops defenders form stopping on a dime, changing directions and then hawking down the Ball Carrier. Out of position defenders will be out of the play. Plain and simple.)

Example: Offensive Code
QB Pump Fakes > Defender not looking at ball > Defender will not bite on the fake
QB Pump Fakes > Defender looking at ball > Defender will bite on the fake depending on his awareness rating

----------
As I discover more, I will continue to update this if it receives positive attention.

[bdoughty]

I would think that one of your goals would be to make some real world examples showing how to create the behavior you are looking for? The majority of us who are not in software design can only go by what little we know about coding and list examples such as yours. You have the unique capability to create these behaviors. I look forward to seeing your future work on the subject.

[lilblaze]

I myself am only a freshman in College. I will not have any hands-on courses until my Sophomore year. So it's very close for me.

On the other hand, people on this website with a good knowledge of Computer Science (Computes Ratings in Madden) and Software Engineering (Game Design/Engine) can also break this down even further. I believe I have supplied the building blocks of how EA works as far as developing Football Games go.

To make a change, you must have knowledge of what needs changing; as changing things blindly can lead to disastrous results.

[JeffHCross]

Could you clarify for me what you're trying to get at here? Based on the title of this thread, I expected something explaining the "EA titles = BAD" conclusion. Seeing that your list has production insight, I expected it would be used as evidence for why their titles are "bad". I don't see either here. There's nothing in your list that says that EA's titles are / are not doing this, except for Madden not being re-written from scratch every year (and, personally, I don't feel like waiting 3, 4 or more years for every Madden game). If nothing else, #7 on your list supports why football titles are so difficult to replicate the real sport.

I haven't read the book you mentioned, but I'm sure it's similar to others that are sitting on my bookshelf. Though since the author was part of the Madden production team for a time, perhaps there are insights there not available elsewhere.

You may also like An Academic's Field Guide to Electronic Arts (https://www.google.com/search?q=An+Academic%27s+Field+Guide+to+Electronic +Arts), which was authored by the late Carneige Mellon professor Randy Pausch (known for "The Last Lecture").


with a good knowledge of Computer Science (Computes Ratings in Madden) and Software Engineering (Game Design/Engine)FWIW, you will find VERY little distinction, outside of (perhaps) academia, between Computer Science and Software Engineering. Obviously you can be a graduate of Computer Science without a focus in SW, but most people equate the two. If you're a CS grad, you're expected to be at least competent as a software engineer.

Thought I'm not sure where you were going with CS = Computes Ratings ...

[lilblaze]

1. Could you clarify for me what you're trying to get at here?

2. There's nothing in your list that says that EA's titles are / are not doing this, except for Madden not being re-written from scratch every year (and, personally, I don't feel like waiting 3, 4 or more years for every Madden game).

3. Thought I'm not sure where you were going with CS = Computes Ratings ...

1. I did explain why EA Football titles are "bad." Much of what I said was implied, assuming it would be understood. Research in Player Ratings hinders game development. Ratings calculated from mass Research, must relate to the "Engine" the football game is running. (Under the assumption that "Engine" by itself works properly in the first place. Making sure that all ratings, high or low don't cause the "Engine" to malfunction.) EA doesn't come out directly and say, "Hey! This is how our Engine works. This how long ratings take to develop. This is how long we have to test our Rated Players on this Engine." When you focus all of your time on single mathematic calculations (Ratings) just to offset a Bad Game Engine, you end up putting out a piece of shit every August. (Assuming the game engine has been the same since 07 (Madden 08), I'd say it's relatively unreliable and outdated.)

Understandable? I hope so.

2. The points you've read are like puzzle pieces. Piece it together and you will see the big picture.

3. Computer Science's role in Player Ratings

Ex.)
QB(1) Accuracy(Total) = 90 (IF)
Chance of Completed Passes to a WR in position to make a play on the ball = 90 (THEN)

The "Engine" has hundreds of animations built into it. EA is delivering games to us that don't even utilize half of those animations.
Since I'm pretty sure that the game consists of IF & THEN Statements (CS), they are linked to the game engine. However many statements are in there, says how many animations we will get. The "Dive Catch" in Madden 12 was implemented as a feature, by pressing Square/X to make a player mindlessly dive and attempt to make a catch they don't even look back to see. I promise it won't be in Madden 13. It will be another string added to the IF & THEN "Catching" category.

[JeffHCross]

1. I did explain why EA Football titles are "bad." Much of what I said was implied, assuming it would be understood.No. You listed "puzzle pieces" and then assumed that someone would put together the puzzle and interpret "BAD". I fail to see the immediate, obvious connection. Furthermore, several of your items aren't even clear on what they're getting at. #7 and #8 read like "AI is hard" and "game events are largely IF/THEN statements". Both of these are true, and tell me nothing about your "Madden is bad" conclusion. If we are to conclude that all of the items you listed are bad, then what are their alternatives?


Research in Player Ratings hinders game development.I truly fail to see how the time that Donny Moore (who, as far as I know, is largely the only person who spends considerable time on player ratings) spends on reviewing players for the sake of accuracy hinders the development of the game. Donny is not a coder (nor a designer, as far as I know).


3. Computer Science's role in Player Ratings

Ex.)
QB(1) Accuracy(Total) = 90 (IF)
Chance of Completed Passes to a WR in position to make a play on the ball = 90 (THEN)Okay, if that is what you meant, why did you separate Computer Science and Software Engineering? That's a dice roll / probability test / simple IF/THEN statement. Statements like that has as much to do with Software Engineering/Development as it does CS.


Since I'm pretty sure that the game consists of IF & THEN StatementsAs do most games with AI or multiple events dependent on dice rolls.

I'm all for a discussion on this, please don't misunderstand, I just was not clear on your original post.

[ram29jackson]

I dont know crap about it and wont knock guys for having jobs in something they went to school for.

But the EA football games seem very one dimensional in some ways compared to other sports games. Yes, they say its hard to do 22 guys with assignments,,,,I dont know ?

why cant ball flight and bounce be a tad more random ? how come the squib kick is more believable in Madden then in NCAA?

[lilblaze]

No. You listed "puzzle pieces" and then assumed that someone would put together the puzzle and interpret "BAD". I fail to see the immediate, obvious connection. Furthermore, several of your items aren't even clear on what they're getting at. #7 and #8 read like "AI is hard" and "game events are largely IF/THEN statements". Both of these are true, and tell me nothing about your "Madden is bad" conclusion. If we are to conclude that all of the items you listed are bad, then what are their alternatives? The title may simply be misleading; a debate is not what I want here. If you have stronger insight than I do, why don't you post plausible alternatives? You seem to understand my article quite well; even though it is somewhat all over the place.

I truly fail to see how the time that Donny Moore (who, as far as I know, is largely the only person who spends considerable time on player ratings) spends on reviewing players for the sake of accuracy hinders the development of the game. Donny is not a coder (nor a designer, as far as I know). This is probably why the game doesn't play accurately to all player ratings.

Okay, if that is what you meant, why did you separate Computer Science and Software Engineering? That's a dice roll / probability test / simple IF/THEN statement. Statements like that has as much to do with Software Engineering/Development as it does CS. Once again, this is not a debate I'm trying to achieve here. My entire article is based on what I read from a game developing book. That's all.

My answers are within the quote.

[skipwondah33]

Like where your head is at regarding obtaining more information into game design.

Software and Video game design was my 2nd major in college..have a few courses to complete still for that one.

[JeffHCross]

My answers are within the quote.Okay, I now understand.

What I primarily struggled with was that you said the games were bad, but there are several items in your list that don't appear to be "bad" to me. Was it the author's viewpoint that Madden was bad, and these were their reasons why? That would be interesting to me.

I believe the game doesn't play "accurately" simply because there is no right or wrong way to assess "accuracy" in this case. "Accurate" would mean that NCAA 08 had Appalachian State beating Michigan, and yet Michigan fans would probably have screamed "BROKEN!" if that had been the case. Now, the easy (and probably most legitimate) argument there is that if Appalachian and Michigan played 100 times, Michigan would likely win the other 99. Which isn't likely to be too far from what NCAA 08, or any NCAA title, would show if you did repeated simulations. There is very little in this game, based on my experience, that is repeatable and testable. Which, I believe, is why those of us on the outside have so little insight on the inner workings of the game. What causes a DB to be in position for an interception? We could probably figure it out, if the game state was repeatably testable. And no, I don't believe Practice Mode counts.

I apologize if I came off harshly, I was trying to make sense of your statements and likely overreacted. I knew you weren't looking for a fight/debate, but I couldn't tell what you were looking for.

While I don't necessarily have alternative suggestions, I did write responses to each bullet last night:
"1. Madden is NOT REDESIGNED every year, nor is its code rewritten.": From scratch? Of course not. Any game with annual, or remotely annual, updates is not going to be re-written from scratch. If you're talking AAA titles, I don't think any sequel has major re-writes in modern game development. Maybe the graphics engine is overhauled. Maybe a new feature is added. But other parts (base logic, networking code, audio, controls) are probably left untouched as much as possible. Though I do think there is more re-design than the Tiburon team is given credit for. Maybe not total code re-write of features, but re-design.

"2. In theory, the best, current football game available on the market today should be NCAA Football 12 and All-Pro Football 2k8." A decent number of people think APF 2k8 is still the best game on the market, though I think that really just comes down to preference. But I have no clue why you say that the majority of time on Madden is research. Research on what? Ratings? I may be wrong, but I don't know that too many people beyond Donny Moore (if any) spend considerable time on research. Donny probably spends more time on roster research than the equivalent guy on NCAA, but that could be largely due to film availability and likeness replication. You said the ratings hinder development. I still fail to see how. Without some kind of internal rating, it is significantly harder to get AI agents to simulate real-life counterparts. If it's acceptable (as it is in some games) for every AI agent to have the same chance of X happening, you don't need ratings. But Peyton Manning is not John Beck. Without ratings (whether visible or not), I'm not sure how that could be represented.

"3. Players greatly enjoy setting up their own strategies, adjusting how the athletes will behave and what roles they will play in a team sport." You mentioned that Madden does this. I agree it could be better -- see Head Coach 09 for an example.

"4. During play, your game will be running a physics engine that DETERMINES THE BEHAVIOR OF MOVING BODIES in the match." Whose game? Is this a theoretical point based on the ideal situation? Sure, a fully functional physics engine is the ideal. Though, technically, a physics engine is still an animation engine.

"5. Even in a highly realistic game, YOUR OBJECTIVE IS TO PROVIDE AN ENJOYABLE EXPERIENCE, not a matchstick simulation of nature." I agree that Madden and NCAA shouldn't provide a absolute simulation of the on-field product (I, for one, do not feel like being chased by a realistic Jason Pierre-Paul or Jason Tuck with a 3 second or less rush), but I don't think "enjoyable experience" and "simulation" are incompatible. I also have no interest in a football game that puts fun to the detriment of simulation. The key, for a simulation sports title, is to find the middle ground.

"6. The ATHLETES in a sports game MUST BEHAVE LIKE HUMANS, and THAT MEANS DELIBERATE, INTELLIGENT ACTION." Agreed.

"7. Athletes must respond intelligently to a number of possible events. A HIGHLY COMPLEX SIMULATION SUCH AS AMERICAN FOOTBALL, with 22 players on the field at a time, presents the opportunity for hundreds of events." Sports AI in general, and football AI specifically (http://aigamedev.com/open/article/football-ai-cheating-machine-learning/), is considered among one of the toughest tasks for artificial intelligence. Agreed. It's actually far beyond "hundreds" if you consider game situation factors (like down, distance and time remaining on the clock).

Here's another link (http://www.sciencedaily.com/releases/2011/08/110818132152.htm) about Football AI. Apparently watching a football game can baffle computers.

"8. Gameplay Code should be designed like a flowchart." Am I to conclude from this that Madden is not designed that way? I expect Madden is designed in a reasonably similar way, and I have seen no evidence to the contrary. Now, some of your flowchart example is an over-simplification of the type of calculation done, but you've got the gist of it.

When I attended the Community Day events last year, I did take advantage of the fact that we were playing on debug consoles, and I turned on the option for some of the debug information to be shown on the screen. Let me just say that I was shocked at the number of variables and factors being tracked for even a simple passing play.

[lilblaze]

@JeffHCross

You are a highly analytical individual. I think you're exploring too much into certain, or most of my statements and/or questions; instead of just viewing/accepting them for what they are. You don't have to dumb anything down, but I'm not looking as much into it as you are. I've accepted the majority of the statements (in the book) for what they are, and included my honest opinions.

About your response to #8: I am 99% sure that Madden, NCAA and 2k Football games are all designed with "Flowchart-Like" code. My deal with Madden is, I feel like they purposely remove strings of code that produce glitches in gameplay (minor or major) -- And they are too lazy to fix them. In one game, I may see the same animation/s 20 times. The prayer catch being one of them. At the rate they're moving, we may only see one or ONE to TWO, NEW and DIFFERENT animated catches each year.

[ram29jackson]

that FTD florest over the shoulder ballet catch is getting old....

[JeffHCross]

I've accepted the majority of the statements (in the book) for what they are, and included my honest opinions.While I'll agree that I'm very over-analytical, I think you raised an important point here as to why you and I are not on the same page. You know which parts are from the book and which parts are your opinion. I don't. So at first I thought it was all your opinion. Then all from the book. And now I know it's somewhere in the middle but not sure where.


In one game, I may see the same animation/s 20 times. The prayer catch being one of them. At the rate they're moving, we may only see one or ONE to TWO, NEW and DIFFERENT animated catches each year.I agree with your observation, that there are few animations. But I don't believe it's because of laziness or removing glitches. Based on some things I've seen evolve in NCAA over the years, I think many animations are replaced from one year to the next. So while we get all "new" animations, the repetition is still there. This may not be true for gameplay animations like catching, but it is something I've observed in other areas (like cutscenes).

[shaunlmason]

Reading this makes my brain hurt. :) I'll be interested to see how your perspective on this topic changes once you start coding. I remember in college thinking, "this is the easiest thing ever, I'll be a billionaire soon!". I didn't realize the work I did in college was relatively trivial until I got a boss to dream up crazy ideas and expect me to deliver them yesterday.

[JeffHCross]

I didn't realize the work I did in college was relatively trivial until I got a boss to dream up crazy ideas and expect me to deliver them yesterday.Yep. Even the hardest assignments in college are cake compared to the systems we're working on in real life.

Though, at the same time, the complexity usually comes over time. Only bitter EA former-fans expect that someone could code an entire game in a year :D